AI Voice in Mexico 2026: From IFT to CRT, LADA Routing, and Biometric SIM

Mexico's telecom regulator was rewritten in 2025: IFT was dissolved, ATDT became the policy authority, and the new CRT (Comisión Reguladora de Telecomunicaciones) took over operational regulation on 21 October 2025. Then on 1 July 2026 every active mobile line — prepaid, postpaid, physical, eSIM — must be linked to a biometric profile. AI voice agents must adapt now.

How telephony works there

Mexican numbering uses 10-digit format with LADA area codes (55 Mexico City, 33 Guadalajara, 81 Monterrey, 222 Puebla, 999 Mérida). Mobile and fixed are integrated under +52, abolishing the historic 044/045 mobile prefix; you dial 10 digits domestically and +52 + 10 internationally. Toll-free is 800 + 7 digits.

The CRT now allocates numbering and oversees licensing; ATDT sets policy. Major operators (Telcel, AT&T Mexico, Movistar) interconnect via class-4 IP and TDM gateways. Twilio Mexico provides DIDs and 800 toll-free, with a regional media edge in São Paulo for ~80 ms RTT to Mexico City and dedicated peering. STIR/SHAKEN-equivalent rules are being drafted; current CLI rules require valid, allocated CLIs and forbid spoofing.

Regulatory landscape

REPEP (Registro Público de Consumidores) is Mexico's equivalent of a do-not-call register, run by PROFECO. Marketing callers must scrub before contacting any consumer and respect opt-out for at least 12 months. The LFPDPPP (Ley Federal de Protección de Datos Personales en Posesión de los Particulares) governs personal-data processing for telephony — including voice — and requires a privacy notice, purpose limitation, and ARCO rights handling. Marketing calls must respect a 09:00-21:30 window in the consumer's local time zone.

The 2025 reform consolidates obligations under the new Ley Federal de Telecomunicaciones y Radiodifusión, with the CRT issuing Lineamientos on caller-ID and CLI authentication, and biometric SIM registration becoming mandatory by 1 July 2026. Failure to comply suspends the line. AI voice agents originating from a non-biometric line will be cut at the network edge.

CallSphere implementation

Explore a live demo and compare current plans to find the right fit for your business.

Build and launch steps

flowchart LR
 A[Twilio MX subaccount] --> B[Provision LADA DID + 800 TF]
 B --> C[Register CLI under tenant RFC]
 C --> D[REPEP nightly scrub]
 D --> E[LFPDPPP consent + AI disclosure]
 E --> F[Time-window 09:00-21:30 enforcement]

1. Create a Twilio Mexican subaccount; pin media to the closest US-MX edge for low RTT.
2. Provision a LADA DID for the tenant's region and an 800 toll-free for inbound.
3. Register the CLI under the tenant's RFC and ensure it is callable end-to-end.
4. Subscribe to REPEP and configure a nightly wash on every outbound list.
5. Capture LFPDPPP consent: timestamp, IP, finalidad, base legal, and store ≥ 5 years.
6. Configure the AI agent to disclose synthetic voice in the first sentence; enforce 09:00-21:30 window.

FAQ

What replaced IFT in Mexico? The Comisión Reguladora de Telecomunicaciones (CRT), which began operations on 21 October 2025; ATDT now sets policy.

Does the biometric-SIM rule apply to enterprise lines? Yes. Every active mobile line — prepaid, postpaid, physical, eSIM — must be biometric-registered by 30 June 2026. New activations have a 30-day window.

Is AI marketing legal in Mexico? Yes, with REPEP scrub, LFPDPPP consent, AI disclosure, and the 09:00-21:30 window.

Do I need a Mexican entity? Twilio MX DIDs can be provisioned without a Mexican entity, but a registered RFC and Mexican operating address improve trust scoring and CRT registration.

Sources

• CRT — Comisión Reguladora de Telecomunicaciones (replaces IFT)
• PROFECO REPEP
• LFPDPPP — Federal Law on Protection of Personal Data
• Yucatán Magazine: 2026 Mobile Registry / Biometric SIM

Explore a live demo and compare current plans to find the right fit for your business.

AI Voice in Mexico 2026: From IFT to CRT, LADA Routing, and Biometric SIM: production view

AI Voice in Mexico 2026: From IFT to CRT, LADA Routing, and Biometric SIM is also a cost-per-conversation problem hiding in plain sight. Once you instrument tokens-in, tokens-out, tool calls, ASR seconds, and TTS seconds against booked-revenue per call, the right tradeoff between Realtime API and an async ASR + LLM + TTS pipeline becomes obvious — and it's almost never the same answer for healthcare as it is for salons.

Serving stack tradeoffs

The big fork is managed (OpenAI Realtime, ElevenLabs Conversational AI) versus self-hosted on GPUs you operate. Managed wins on cold-start, model freshness, and zero-ops; self-hosted wins on unit economics past a certain conversation volume and on data residency for regulated verticals. CallSphere runs hybrid: Realtime for live calls, self-hosted Whisper + a hosted LLM for async, both routed through a Go gateway that enforces per-tenant rate limits.

Latency budgets are non-negotiable on voice. End-to-end target is sub-800ms ASR-to-first-token and sub-1.4s first-audio-out; anything beyond that and turn-taking feels stilted. GPU residency in the same region as your TURN servers matters more than choosing a slightly bigger model.

Observability is the unglamorous backbone — every conversation produces logs, traces, sentiment scoring, and cost attribution piped to a per-tenant dashboard. HIPAA aligned isolation keeps healthcare traffic separated from salon traffic at the storage layer, not just the API.

FAQ

What's the right way to scope the proof-of-concept? Setup runs 24 hours, the trial is 7 days with no credit card, and pricing tiers are $19, $49, $99, and $149 — so a vertical-specific pilot is a same-week decision, not a quarterly project. For a topic like "AI Voice in Mexico 2026: From IFT to CRT, LADA Routing, and Biometric SIM", that means you're not starting from scratch — you're configuring an agent template that's already been hardened across thousands of conversations.

How do you handle compliance and data isolation? Day one is integration mapping (scheduler, CRM, messaging) and prompt tuning against your top 20 real call transcripts. Day two through five is shadow-mode running, where the agent transcribes and recommends but a human still answers, so you can compare side-by-side. Go-live is the moment your eval pass-rate clears your internal bar.

When does it make sense to switch from a managed model to a self-hosted one? The honest answer: it scales until your tool catalog gets stale. The agent is only as good as the integrations it can actually call, so the operational discipline is keeping schemas, webhooks, and fallback paths green. The platform handles the rest — observability, retries, multi-region routing — without your team owning the GPU layer.

Talk to us

Explore a live demo and compare current plans to find the right fit for your business.